关键词: 智能化检测, 图像处理智能算法, 语音识别, 异常状态自动标注, 语音报警

Abstract: The proposed technical optimization is used to solve such problems as "uncertainty", "high intensity" and "low efficiency" in the traditional operation mode of the current station train receiving and departure operations and railway cargo inspection operations relying on manual whole-process monitoring and subjective judgment. The various collected data at the detection points are effectively integrated, such technologies as intelligent image recognition, speech recognition, thermal imaging, machine learning algorithm and system integration are comprehensively utilized, and the train operation and freight train loading states are intelligently detected in the multiple-direction, multiple-angle and multiple-type manner, to achieve the automatic recognition and labeling of and voice alarming for various abnormal train states. This study has great significance for the comprehensive improvement of the quality and efficiency of train receiving and departure operations and freight inspection operation, and also the guarantee of the safety of railway transportation.

Key words: intelligent detection, intelligent algorithm of image processing, speech recognition, automatic labeling of abnormal state, speech alarming